Synthesis of failure rates for space station external orbital replaceable units

In 1990 NASA commissioned a study of the maintenance requirements of external orbital replaceable units (ORUs) of Space Station Freedom, which are largely determined by ORU failure rates. Freedom ORUs were still in preliminary design and had no operational or unit-level test experience, so it was necessary to use conventional design-phase reliability engineering techniques to estimate their reliability. Because of widespread doubts about the validity of these methods in the space flight community and the importance of the external maintenance burden to the practicability of the Station concept, NASA needed a way to independently confirm the validity and credibilty of the designersʹ failure rate estimates. The novel and effective failure rate synthesis approach which was developed to fill this need had the following elements: (1) enforcing standard, consistent, technically sound reliability prediction methods over all external ORUs and ORU suppliers; (2) checking the resulting failure rates against the range of rates achievable with current technology, based on a synthesis of expert engineering judgement; and (3) also checking the designersʹ rate against the failure experience of the analogous items on existing spacecraft, with appropriate corrections for scale and application differences. mediate result of the failure rate study was to show that the designersʹ failure rates derived by conventional reliability engineering techniques, when held to a consistent standard methodology, were consistent with both of the credibility checks and hence could be relied upon for predicting the Station-level external maintenance rate. More broadly, the consistency among the three failure rate estimating methods indicates that the synthesis of expert judgement based on available technology is a valid, cost-efficient way to estimate the bounds of achievable reliability for equipment that not only lacks operating experience, but is still in the preliminary or even the conceptual stage of development. This provides a new engineering tool for evaluating the reliability, availability, maintainability, and supportability of systems in the very earliest phases of development, when changes can be made most economically.